172 JOURNAL OF THE ROYAL HORTICULTURAL SOCIETY. 



parthenogenetic generations produced from female forms by 

 viviparous " budding " or the laying of unfertilised eggs. The 

 host on which the sexual generation is found, which probably 

 represents the primitive food-plant of the species, is spoken of as 

 the "primary host" or primary food-plant," the other being 

 the "intermediate food-plant." In Chermes Ahietis-laricis (as 

 these insects should henceforth be called), the Spruce, upon which 

 the sexually produced egg is always laid, is the primary food- 

 plant, the Larch is the intermediate food-plant. 



The cycle given for these insects by Dreyfus is as follows : — 



First Year. 



Generation I. passes the winter on the Spruce as Clier^nes 

 Abietis, and there lays eggs. 



Generation II. develops in the Spruce-galls, and emerges 

 as the winged forms of Abietis in August. Part of these emi- 

 grate to the Larch, and, being now known as Laricis, lay eggs 

 on its needles. 



Generation III. hatches on the Larch, and passes the winter 

 under bark-scales, &c. 



Second Year. 



From the eggs of Generation III. appear Generation IV. 

 (the woolly form of C. Laricis), which acquire wings at the end 

 of May, and mostly return to the Spruce. There they are known 

 as G. obtectus* and their eggs produce 



Generation V., the bisexual, consisting of true males and 

 females. From their sexually produced eggs develop gradually 

 from July to September the " Stem-mothers," which live through 

 the winter and correspond to Generation I. of the cycle. 



Blochmann, who contests with Dreyfus the priority of these 

 discoveries — a dispute into which it is not necessary to enter — 

 gives a life-history differing in the presence of a second gall- 

 inhabiting generation, and an additional one passing the winter 

 on the Spruce. These two generations will be interpolated 

 between I. and II. of the cycle just quoted, and the whole series 

 will extend over three years. The exact number of generations 

 is, however, not important in practice, and it is sufficient 



* This form appears to be as yet unrecorded from Great Britain, and 

 its identification with the remigraut form of C. Laricis is disputed by 

 Blochmann. 



